Exposed: 90% of SIU referrals fail to lead to NPA prosecutions
The recent briefing revealed a shocking 90% failure rate of SIU referrals to the NPA, igniting public outrage over the state’s corruption crisis and the erosion of ethical standards.
A recent briefing between Parliament’s Standing Committee on Public Accounts, the National Prosecuting Authority, the Special Investigating Unit and the Hawks unveiled a critical issue: only 10% of the 5,994 corruption referrals since 2011 have resulted in registered criminal dockets. Committee members expressed outrage at the slow pace of prosecutions and the erosion of public trust in state institutions.
The NPA defended its slow progress, citing the consolidation of multiple referrals into a single investigation, as well as the difficulty in securing forensic auditors, witnesses, and documentation. Governance expert Sandile Swana accused the system of being captured and misdirected, pointing to the competence of prosecutors like Andy Mothibi, who were once SIU heads.
The Madlanga Commission also revealed that some "well prepared cases" were ignored, casting doubt on the credibility of investigators. The NPA spokesperson Kaizer Kganyago acknowledged that not all referrals lead to criminal investigations, but claimed that the statistics would never match due to the merging of multiple referrals.
The commission's scrutiny of the NPA's anti-corruption arm, IDAC, led to the resignation of IDAC head Andrea Johnson and the withdrawal of charges against top officers. Former prosecutors also faced corruption charges. Analyst Zakhele Ndlovu highlighted the government's lack of capacity and political will to combat corruption, while Professor Andre Duvenhage warned of political interference within the NPA, making prosecutions and investigations uneven and selective.
Out of the 304 matters referred to the NPA by the SIU, only 81 convictions, including 17 custodial sentences ranging from 8 to 15 years, have been secured.
Written by urgent.news from IOL's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.